Is Traffic Still an Important Emitter of Monoaromatic Organic

Nov 22, 2017 - Aerosol Health Effects from Molecular to Global Scales. Environmental Science & Technology. Shiraiwa, Ueda, Pozzer, Lammel, Kampf, Fush...
0 downloads 10 Views 2MB Size
Subscriber access provided by READING UNIV

Article

Is traffic still an important emitter of mono-aromatic organic compounds in European urban areas? Agnès BORBON, Anne Boynard, Therese Salameh, Alexia Baudic, Valerie Gros, Julie Gauduin, Olivier Perrussel, and Cyril Pallares Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b01408 • Publication Date (Web): 22 Nov 2017 Downloaded from http://pubs.acs.org on November 23, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 28

Environmental Science & Technology

Is traffic still an important emitter of mono-aromatic organic compounds in European urban areas?

1 2

Agnès Borbon1*, Anne Boynard2, Thérèse Salameh3ǂ, Alexia Baudic4,5 Valérie Gros4, Julie Gauduin5, Olivier Perrussel5, Cyril Pallares6

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

1. Université Clermont Auvergne, CNRS, LaMP, F-63000 Clermont-Ferrand, France 2. LATMOS/IPSL, UPMC Univ. Paris 06 Sorbonne Universités, UVSQ, CNRS, Paris, France 3. Laboratoire Inter-Universitaire des Systèmes Atmosphériques (LISA), UMR CNRS 7583, Université Paris Est Créteil et Université Paris Diderot, Institut Pierre Simon Laplace, Créteil, France 4. Laboratoire des Sciences du Climat et de l’Environnement (LSCE), UMR CNRS 8212, Commissariat à l’Energie Atomique et aux énergies alternatives, Université Versailles Saint-Quentin, Institut Pierre-Simon Laplace, Gif sur Yvette, France. 5. AIRPARIF, Association Agréée de Surveillance de la Qualité de l’Air en Île-deFrance, 75004 Paris, France 6. Agence de surveillance de la qualité de l’air (ASPA), F-67300, Schiltigheim, France

18 19

ǂ

20 21

* corresponding author [email protected]

now at : IMT Lille Douai, Univ. Lille, SAGE - Département Sciences de l'Atmosphère et Génie de l'Environnement, 59000 Lille, France

1

ACS Paragon Plus Environment

Environmental Science & Technology

Page 2 of 28

22

Abstract

23

Trends of long-term observations and emission inventories suggest that traffic emissions will

24

no longer dominate the concentrations of mono-aromatic compounds (i.e. TEX - Toluene,

25

Xylenes and Ethylbenzene) in European urban areas. But the split limit between traffic and

26

other emission sector contributions such as solvent use remains tenuous. Here long-term

27

observations of an extensive set of hydrocarbons, including TEX, at traffic and urban

28

background sites in London, Paris and Strasbourg were combined to estimate the relative

29

importance of traffic emissions on TEX in every city. When analyzing the urban enhancement

30

emission ratios of TEX-to-benzene on a seasonal basis, two potential source signatures other

31

than traffic could be differentiated in all cities (1) summertime evaporation from fuel and/or

32

solvent and (2) wintertime domestic heating. However, traffic emissions still unambiguously

33

dominate the concentration levels of TEX in every city despite the reduction of their

34

emissions at exhaust pipe over the last two decades. Traffic explains between 60% and 96%

35

(at ±20%) of TEX levels while it is less clear for xylenes at some locations. Our results

36

provide a basis to evaluate regional emission inventories. The method is applicable at any

37

urban area where speciated hydrocarbon monitoring is available.

38 39

Toc Art 2

ACS Paragon Plus Environment

Page 3 of 28

40

41

Environmental Science & Technology

Key points:



Emission inventories and multi-year trends of long-term observations raise the

42

question of the impact of traffic emissions on TEX concentration levels in European

43

urban areas

44



One decade of hourly observations of hydrocarbons at traffic and urban stations in

45

Paris, London and Strasbourg is used to estimate the traffic fraction of TEX

46

concentrations

47



emission composition and additional wintertime domestic combustion for benzene

48

49

The seasonal signature of TEX sources is rather due to seasonal changes in traffic



It is demonstrated that traffic emissions still account for more than 60% of most TEX

50

concentrations regardless of the season while is less clear for xylenes at some urban

51

locations

52

3

ACS Paragon Plus Environment

Environmental Science & Technology

Page 4 of 28

53

Introduction

54

BTEX (Benzene, Toluene, Ethylbenzene and (m+p) and o-Xylenes), a sub-group of Volatile

55

Organic Compounds (VOC), still constitute a significant fraction of VOCs in urban

56

atmospheres up to 47% in volume of total Non-Methane Hydrocarbons (NMHC) after C2-

57

C9 alkanes1;

58

emission sources (transport, storage and distribution of fossil fuels, industrial and residential

59

combustion, processes) and solvent use6. For only benzene, which is known to be

60

carcinogenic7, the European air quality objective for health protection of 5 µg.m-3 (1.57 ppb

61

STP) per year is still exceeded in the center of Paris close to major roadways8. Apart from

62

their abundance and adverse health effects, BTEX have significant implications on

63

atmospheric chemistry. Together with alkenes, they show the highest ozone-forming

64

potential9,10. They have therefore been considered as key target compounds by ozone policy

65

reduction measures, especially in urban areas where ozone is usually VOC-limited11. BTEX

66

oxidation also contributes to the formation of Secondary Organic Aerosols (SOA). While high

67

uncertainties persist on SOA yields depending on smog-chamber experimental conditions,

68

BTEX show SOA yields as significant as the one of α-pinene, the most emitted and reactive

69

biogenic precursor being currently the dominant SOA source at the global scale: yield values

70

range from 1181 (benzene) to 1578 µg.m-3.ppm-1 (xylenes), respectively12,13 against

71

1826 µg.m-3.ppm-1 for α-pinene14. Recent results have suggested that BTEX could explain

72

34% of SOA production in summer15 in Paris megacity whereas such a contribution would be

73

lower when taking into account a larger set of VOC of intermediate volatility (IVOC) in Los

74

Angeles16. Like terpenes, oxidation of mono-aromatics would also lead to the rapid formation

75

of highly oxygenated molecules as recently suggested by Molteni and coworkers17.

2; 3; 4, 5

. BTEX are released into the atmosphere by either fossil fuel related

4

ACS Paragon Plus Environment

Page 5 of 28

Environmental Science & Technology

76

VOC emission reduction measures, such as emission reduction at vehicle exhaust pipe, have

77

successfully reduced their anthropogenic concentrations in cities like London18, Paris and

78

other medium-sized cities in France6, Los Angeles19 and at rural stations

79

trends of monthly average concentrations of toluene (Figure 1) and other aromatics (Figures

80

S1, S2, S3 in Supplement Information) at both urban background and traffic sites in three

81

European metropolitan areas (London, Paris and Strasbourg) support these findings.

82

Regardless of the city, all concentrations have consistently shown a clear decline at traffic

83

sites since 2000, especially in the first half of the decade. However, aromatic levels at urban

84

background sites have stayed almost constant since 2008 and concentrations at traffic sites

85

tend towards the urban background levels. This raises the question whether traffic emissions

86

still dominate TEX emissions. Up-to-date emission inventories in all three cities (see pie-

87

charts on the right hand panel of Figure 1 and Figures S1 to S3 in SI) show various features. It

88

should be noted that no local emission inventory is available for London megacity and the

89

emission distribution is adapted from the National Atmospheric Emission Inventory in UK

90

(NAEI) for the year 2000 (D. Derwent and S. Beevers’ personal communication). For toluene,

91

the road-transport sector accounts for half of its emissions in London and Paris and only 10%

92

in Strasbourg. For xylenes (Figures S1 and S2), the road-transport contribution drops down to

93

4%-33% except for o-xylene in London (57%). The solvent use and/or residential sector

94

appear as the other major source explaining between 28% and 56% of toluene emissions and

95

44 to 77% especially for xylenes. Only ethylbenzene would be still largely emitted by the

96

road-transport sector (44 – 68 %) (see Figure S3). Regional emission inventories for two

97

European cities, Paris and Zurich in 2005, also pointed to the use of solvents from all source

98

categories as the new dominant source of total VOCs at the urban scale. Finally, the

99

homogeneity of aromatic multi-year trends depicted in observations is not reflected in

100

regional emission inventories. However emission inventories for VOCs still show high

20,6

. The multi-year

5

ACS Paragon Plus Environment

Environmental Science & Technology

21

Page 6 of 28

101

uncertainties up to a factor of 4

and observations provide useful constraints for their

102

evaluation22;

103

usually point to the contribution of various sources to total measured VOC ambient levels26; 27;

104

24

105

solvent use-related activities and wood combustion for benzene26,

106

supporting source-receptor models directly imply an estimation of source/factor contributions

107

to total measured VOC concentrations rather than individual species. Towards the complexity

108

of TEX emission sources, there is still no consensus on the importance of BTEX source and

109

for traffic in particular. Moreover, the seasonality of emission changes has been poorly

110

studied for BTEX.

111

In this study, we combine long-term and continuous in-situ measurements of NMHC,

112

including BTEX, carried out in European urban areas by air quality monitoring network

113

programs over the past decade to investigate the importance of traffic emissions on mono-

114

aromatic levels at the urban scale. Due to its stringent regulation in fossil fuel and solvent

115

formulation compared to other mono-aromatic compounds, benzene will be treated here as a

116

combustion source tracer rather than a target species for the quantification of traffic

117

contribution. This study provides a scientific basis to evaluate the accuracy of regional

118

emission inventories.

119

Experimental

120

Data analysis mostly relies on continuous hourly observations of a large set of NMHC,

121

including BTEX, performed by European air quality networks in the megacities of London

122

(UK by DEFRA), Paris (France by AIRPARIF) and Strasbourg (Eastern France by ASPA).

123

The chosen period for data analysis covers the years 2008 to 2013. Special care was put on

124

data quality checks. The description of quality assurance tests is provided in Boynard et al.

23 24; 21;25

. Source-receptor models like PMF (Positive Matrix Factorization)

. BTEX usually show significant contributions in the source profiles of vehicle exhaust, 27, 24, 28

. Mathematics

6

ACS Paragon Plus Environment

Page 7 of 28

Environmental Science & Technology

125

(2014)2 and Waked et al. (2016) 6 for French data and Dollard et al. (2007)29 for UK data. The

126

purpose here is to evaluate the importance of traffic emissions relative to other anthropogenic

127

emissions. Data analysis has therefore combined observations collected at both traffic and

128

urban background sites in each city to provide perspectives from contrasted urban situations

129

non-influenced by industrial emissions. Traffic sites represent those locations under the direct

130

influence of nearby traffic emissions, usually a roadway tolerating traffic greater than 10,000

131

vehicles per day. Urban background sites are under the influence of multi-emitters, at a

132

distance of major traffic lanes of 40 m on average and are more likely representative of the

133

urban mixing of all anthropogenic emissions. Details on the database are reported in Table 1.

134

A description of the sites, meteorology and analytical devices is provided in the Supplement

135

Information (pages S5-S6). Note that the Paris data has been completed by tunnel

136

measurements performed during the PREQUALIF project 28. The number of missing data is

137

usually lower than 25% over the six-year period.

138

Results and discussion

139

Concentration levels of BTEX

140

Average concentration levels of mono-aromatics including benzene in the three metropolitan

141

areas are reported in Table S1 (2nd to 5th column) in the Supplement Information for the year

142

2011. BTEX concentration levels range between 0.03 ppb (ethylbenzene at Strasbourg Ouest

143

and 1,74 ppb (toluene at Paris-Auteuil), i.e. two orders of magnitude. Concentration levels are

144

2 to 4 times higher at the traffic sites compared to urban background sites. Concentrations of

145

benzene are quite similar between the three cities at ±0.040 ppb at traffic sites and ±0.070 ppb

146

at urban background sites. This is consistent with its stringent regulation policy at European

147

level to achieve the air quality standard of 5 µg.m-3 per year by, for instance, limiting its

148

presence at 1% vol. in fuels (Directives 98/70/EC replace by 2009/30/EC and 2003/17/EC). 7

ACS Paragon Plus Environment

Environmental Science & Technology

Page 8 of 28

149

On the opposite, the sum of BTEX load is higher in Paris by 1.4 at traffic sites to 2.4 at urban

150

background sites while concentration levels in London and Strasbourg are surprisingly quite

151

comparable. However, the atmospheric composition in BTEX is very similar between near-

152

traffic and urban background atmospheres. Together with (m+p)-xylenes, toluene (which is

153

the dominant species among BTEX) explains 70% of BTEX load. This suggests that BTEX

154

concentrations are driven by emissions with a non-visible effect of photochemical processing

155

at the background sites. For the latter we would have expected an enrichment of air mass

156

composition in longer-lived mono-aromatics like benzene given that BTEX reactivity towards

157

OH radical spans one order of magnitude30. For benzene, a domestic heating source is highly

158

suspected, especially in Strasbourg (see following sections) and as recently shown by Baudic

159

and co-workers28.

160

Seasonal source signature on TEX

161

The magnitude of TEX concentrations is affected by the proximity of emission sources, the

162

extent of photochemical processes and boundary layer conditions. To minimize those effects,

163

we use the TEX-to-benzene enhancement emission ratio (ER = ∆[TEX]/∆[benzene]) at urban

164

background sites to identify the TEX source signature on a seasonal basis in order to index

165

seasonal changes in urban enhancement ratios to potential changes in the nature and intensity

166

of their sources. The TEX-to-benzene enhancement ratio equals the slope of a linear 2-sided

167

fit of a TEX-to-benzene correlation plot. An illustration of summertime (July) and wintertime

168

(January) scatterplots is provided in Figure 2a over the six-year period at the urban

169

background sites. Urban enhancement ER of toluene-to-benzene representative of traffic,

170

domestic wood combustion and gasoline evaporation emissions are also reported to provide a

171

comparison basis, which is rather qualitative than quantitative regarding the numerous factors

172

modifying the emission composition. These three emission categories have been usually 8

ACS Paragon Plus Environment

Page 9 of 28

Environmental Science & Technology

173

identified as TEX major sources in urban areas21;

24; 28

174

values of ER is described in the Supplement Information (pages S7-S8). Emissions from

175

solvent use cannot be excluded but representative ER relative to benzene can hardly be

176

determined. The main reason is due to the diversity of solvents. Speciated emission profiles

177

usually show great variability with, for instance, toluene-to-benzene ratios spanning two

178

orders of magnitude31. Moreover, benzene, due to its toxic and carcinogen nature, has been

179

strongly limited in solvent formulations. Current standards in Europe have established limits

180

in benzene at 0.1% in cleaning products.

181

Remarkably, scatterplots show similar multiform patterns in all urban areas (Figure 2a). The

182

scatterplots lie within a domain delimited by the evaporative source ER in its upper part and

183

the domestic wood combustion ER in its lower part. Within these limits, the wintertime and

184

summertime scatterplots roughly split along the traffic ER line, suggesting the additional

185

influence of sources during both seasons despite expected seasonal changes in the

186

composition of traffic emission. In summer, the scatterplot equals (Strasbourg) or is above the

187

traffic ER line (Paris and London) and tends to the one of evaporative emissions. This would

188

suggest the presence of fugitive evaporative processes in addition to traffic. In winter, the

189

scatterplot lies between the traffic ER in its upper part and the domestic wood combustion ER

190

in its lower part. This would suggest an additional source of wood combustion for benzene27,

191

28

192

benzene mixing ratios than the ones of toluene especially in Strasbourg, a city submitted to

193

well-marked continental climatic conditions. For Strasbourg this is consistent with the

194

emission inventory (Figure 1). Note also that the multiform pattern is temperature-dependent

195

as seen in Paris and Strasbourg when color-coded with temperature (Figure S4 in Supplement

196

Information). Toluene mixing ratios stay rather constant between both seasons. One cannot

. The determination of the reported

. Note that the observed multiform patterns are rather driven by the seasonal variability of

9

ACS Paragon Plus Environment

Environmental Science & Technology

Page 10 of 28

197

exclude that the correlation slope can be affected by photochemical processing for most

198

reactive VOCs21,22 especially in summer. When applying the nighttime and daytime filters

199

values following Borbon et al. (2013)30, no significant difference is depicted suggesting that

200

emissions mainly control TEX concentrations (not shown here). It is consistent with

201

observations during MEGAPOLI-2010

202

concentrations was only visible in the Paris plume explored by the French ATR-42 aircraft

203

during the MEGAPOLI campaign in the summer 2009 at a distance above 50 km, at least,

204

from the urban center30..

205

The correlation slopes have been determined on a monthly basis at traffic and urban

206

background sites. The toluene-to-benzene ERs are reported in Figure 2b for Paris and

207

Strasbourg. As expected, a clear seasonal variability is depicted on ER time series with a

208

higher variability at urban background sites. However, both cities show an opposite pattern. In

209

Paris, ERs are similar in winter and the urban background ER significantly increases in

210

summer compared to the traffic ER. In Strasbourg, the summertime ERs are similar while the

211

urban background ER decreases compared to the traffic ER in winter. Such temporal and

212

spatial variability also suggests the potential gaining importance of summertime sources other

213

than traffic as in Paris (evaporation) and the potential gaining importance of wintertime

214

sources other than traffic as in Strasbourg (residential wood combustion).

215

Finally, seasonal scatterplots of TEX mixing ratios versus benzene mixing ratios and time

216

series of TEX-to-benzene ER suggest a seasonal change in traffic emission composition and

217

the potential influence of sources other than traffic to BTEX concentrations. The importance

218

of these additional source contributions is going to be evaluated in the next section.

219

Contribution of traffic emissions to TEX concentrations

30.

in Paris. The effect of photochemistry on TEX

10

ACS Paragon Plus Environment

Page 11 of 28

Environmental Science & Technology

220

The method consists in estimating the fraction of each TEX originating from traffic emissions

221

at the urban background sites which are representative of the urban emission mixing. This

222

fraction is calculated by dividing the TEX concentration originating from traffic emissions to

223

the measured concentration. The TEX concentration originating from traffic emissions

224

([TEX]traffic) is expressed as follows:

225

[TEX]traffic = ([tracer]traffic – [tracer]bckgd )× ERtraffic

(equation 1)

226

where [tracer]traffic is the concentration of a traffic tracer subtracted by its regional background

227

concentration ([tracer]bckgd) which could be significant for longer-lived compounds like

228

benzene28 and ERtraffic is the urban enhancement emission ratio from traffic between TEX and

229

the traffic tracer. This equation is derived from the one developed by Borbon et al. (2003)21.

230

Acetylene has been commonly used as a traffic exhaust tracer21, 32. However, a non-exhaust

231

combustion source cannot be excluded especially during the winter season regarding domestic

232

heating as for benzene

233

significant contributions in the wood burning factor solution calculated by PMF 27, 28. To take

234

into account potential additional sources to traffic and seasonal changes in their relative

235

importance, five hydrocarbon tracers have been selected in order to derive an uncertainty of

236

the TEX traffic fraction. These tracers include: four combustion products/tracers (ethylene,

237

propene, acetylene and benzene) for which a wintertime domestic heating combustion source

238

cannot be excluded and an unburned-fuel VOC (isopentane) for which a gasoline evaporation

239

source has been demonstrated in summer21, 3. In addition, source attributions for the five

240

tracers by the AIRPARIF (Paris) and NAEI (UK) emission inventories are reported in the

241

Supplement information (figure S8).

28

. In past studies, acetylene and ethylene have already shown

11

ACS Paragon Plus Environment

Environmental Science & Technology

Page 12 of 28

242

ERtraffic is determined from the slope of a two-sided linear regression fit between TEX at

243

traffic rush hours and the tracer measured at the traffic sites. For Paris, the determination of

244

ERtraffic for TEX combines ERs relative to benzene at the traffic site (Auteuil) and ER of

245

benzene relative to other tracers from the Prequalif tunnel experiment (see 6th column of

246

Table S1). While the ERtraffic from the PREQUALIF experiment was only determined during

247

the fall period (September-October 2012), it is representative of the average of expected

248

monthly ERs, the maxima and minima usually being observed in summer and winter (see

249

discussion below). While the contribution at the traffic site of additional sources can be

250

neglected, some seasonal changes in vehicle fuel composition can be expected as a result of

251

seasonal changes in fuel composition regulation and temperature-dependent evaporative

252

emissions. ERtraffic was therefore determined on a monthly basis. ERtraffic relative to the five

253

tracers values are reported in Table S1 at the three traffic sites. The monthly variability of

254

ERtraffic is significant showing minimum and maximum values in winter and summer,

255

respectively. Minimum values from near-traffic measurements found in the literature are also

256

reported for comparison. All TEX are highly correlated with the five tracers at the traffic sites

257

with an R2 higher than 0.80 and a narrow distribution along the regression slope as an

258

evidence of the traffic origin for TEX (see the example in figure S7 for toluene versus

259

benzene at traffic sites). Values of TEX-to-benzene ER are in the same range among the three

260

cities. It should be noted that ER in Paris and London are in the upper range compared to

261

Strasbourg (only toluene-to-benzene ER reported) and other ER derived from tunnel

262

measurements in Europe. This enrichment has been first pointed out by Borbon et al. (2013)

263

and would be due to two-wheeler powered vehicles (Salameh et al., to be submitted in 2017).

264

Note that the ER relative to benzene is quite similar and the greatest variability is usually

265

observed for toluene and (m+p)-xylenes. Values of ER of TEX relative to other tracers are

266

also consistent between Paris and London. 12

ACS Paragon Plus Environment

Page 13 of 28

Environmental Science & Technology

267

[tracer]bckgd is determined on an hourly basis and is derived from a linear interpolation of

268

minimum concentrations observed over a 4-day period between 2008 and 2013. The method

269

was validated by comparing the time series of interpolated background concentrations to the

270

ones observed at European rural sites in the frame of the EMEP monitoring network and

271

extracted from the EBAS database (http://ebas.nilu.no). VOC data collected twice a week is

272

available from 2008 to 2013 during a 4-hour sampling by canisters. For France, two rural sites

273

have been selected: La Tardière (LT, Western France) and Peyrusse-Vieille (PV, Southern

274

France)20. As an example, a comparison is provided on Figures S5 and S6 in the Supplement

275

Information for the five tracers for Paris and Strasbourg. Except for some periods when local

276

influences at the rural areas cannot be excluded, especially under a continental influence and

277

analytical uncertainties, the time series of interpolated urban background minimum

278

concentrations are consistent with the ones at rural background sites. Median ratios between

279

calculated backgrounds and the ones at Peyrusse-Vieille and La Tardière equal 1.25 and 0.76,

280

respectively.

281

In Figure 3, we have compared the average mixing ratios derived from the traffic

282

(ERtrafic×[tracer]) and background ([TEX]bkgd) terms to the observed mixing ratios in the case

283

of toluene at the three urban background sites. TEX background concentration was determined

284

following the 4-day minima interpolation method described previously for the five tracers. The

285

summertime average contributions of traffic, background and non-traffic terms are illustrated

286

by pie charts and the values of annual traffic contributions to measured concentrations are

287

reported below.

288

Regardless of the location, toluene concentrations show the typical seasonal variability of

289

anthropogenic VOC with maximum concentrations in winter and minimum concentrations in

290

summer. On an annual basis and regardless of the location, traffic emissions unambiguously 13

ACS Paragon Plus Environment

Environmental Science & Technology

Page 14 of 28

291

dominate toluene mixing ratios by explaining more than 70% of the observed concentrations.

292

Given the use of five tracers, the standard deviation of traffic fraction was estimated to be

293

within a ± 13-24 % range. The background contribution is significant with an annual average

294

of 7% in Strasbourg, 16% in London and 20% in Paris. Contrary to other mono-aromatics, a

295

high background level was found for toluene in Paris exceeding by almost 0.40 ppb the rural

296

ones. We suspect the background concentration being overestimated by our method. We

297

cannot exclude that the estimated background include a non-traffic source term. However,

298

when substracting the background term, traffic emissions explains more than 90% of toluene

299

concentrations. In summer, the traffic contribution remains higher than 58% (Paris, 2013) and

300

higher than 78% without accounting for the background term. This is not significantly different

301

from the annual contribution. The expected additional contribution of summertime evaporative

302

sources (like gasoline evaporation from Figure 2) is not depicted. Firstly, our apportionment

303

method (see equation 1) takes into account the seasonal variability of traffic emission

304

composition by the determination and the use of monthly ER values (Table S1). Secondly, this

305

suggests that the shape of the scatterplots reported in Figure 2a is driven rather by the seasonal

306

change of traffic emission composition and additional non-traffic sources in winter like

307

domestic combustion for the tracers. It should be noted that the sum of the traffic and

308

background terms in winter is usually higher than observed concentrations, indicating that

309

combustion sources other than traffic contribute to the tracer mixing ratios. The strongest

310

evidence is seen at the continental background site of Strasbourg. When only considering

311

isopentane as a traffic tracer in winter in Strasbourg, the traffic contribution falls by 20%

312

compared to the one by combustion tracers but remains above a 50%-contribution (not shown

313

here). However, our results suggest that the traffic source contribution has decreased by 31%

314

in Paris over the last six years. This trend is not depicted in Strasbourg and London. Further

315

analysis for the years following 2013 should be performed to confirm this trend. On an annual 14

ACS Paragon Plus Environment

Page 15 of 28

Environmental Science & Technology

316

basis, these results are consistent with emission inventories in the megacities of Paris and

317

London (Figure 1) even if the observed traffic contribution is higher. For Strasbourg, the

318

emission inventory only allocates 10% of toluene emissions to traffic (Figure 1) which would

319

be largely underestimated as suggested by our results.

320

Annual values and summertime pie charts of the traffic contribution are compiled for all other

321

mono-aromatic compounds in Table 2. The values of summertime traffic contribution are

322

usually within the standard deviation of the annual ones. While the split limit between traffic

323

and non-traffic contribution is more tenuous for >C7-aromatics (e.g. (m+p)-xylenes in

324

Strasbourg), traffic emissions would still dominate the concentrations of >C7-aromatics. The

325

summertime (and annual) contribution of traffic is usually higher than 60% and up to 96% over

326

the last six years in the three cities. One exception is the 2013-contribution with a 43%-value

327

to o-xylene concentrations in London and a 46%-contribution for (m+p)-xylenes in Strasbourg

328

in 2009. In London, the summertime traffic contribution to o-xylene concentrations has

329

decreased from 86% to 43%. The decreasing trend for (m+p)-xylenes in London is surprisingly

330

not as clear as for o-xylene. This finding is consistent with the observed change at Eltham in

331

the o-xylene-to-(m+p)-xylenes when looking at the shape of the scatterplot (not shown here)

332

Further analysis should be performed after 2013 to confirm. These results are qualitatively

333

consistent with emission inventories for ethylbenzene but usually disagree for xylenes in the

334

French cities of Paris and Strasbourg, for which the solvent use dominates by far their

335

emissions (Figure 1).

336

In Paris, the traffic contribution for the year 2010 is compared to the one extracted from a PMF

337

source-receptor study by Baudic et al.28 for the same reference year. Baudic and co-workers28

338

performed a PMF analysis on a one-year hourly VOC dataset, including the NMHC data used

339

here. They found an optimized solution of 6 factors, including a Motor Vehicle Exhaust, an 15

ACS Paragon Plus Environment

Environmental Science & Technology

Page 16 of 28

340

Evaporative Sources and a Solvent Use factor. The Evaporative Source factor is related to

341

gasoline evaporation as well as other fossil fuel storage and distribution activities. The

342

contribution of BTEX to this factor being less than 1% , the following discussion will only

343

focus on the two other factors. From the fraction of each TEX in the factor profiles and the

344

modelled concentration of each factor, the modelled contribution of the Motor Vehicle Exhaust

345

(MVE) Factor was calculated to TEX concentrations predicted by the PMF model. This

346

comparison should be interpreted with caution since (i) the MVE contribution is calculated on

347

TEX predicted concentrations (ii) PMF source profiles are assumed to be constant during the

348

whole year in the PMF calculation (iii) assumptions regarding factor identification. For the

349

latter, factor identification is consolidated by comparing to near-source measurement profiles

350

and independent source tracer such as BCff (Black Carbon from fossil fuel combustion) for the

351

MVE factor. Regarding item (i), note that observed concentrations are 30% higher than the

352

ones predicted by the PMF 6-factor solutions for toluene and higher than 20% for C8-

353

aromatics. Regarding item (ii) ER calculations have shown a clear seasonal variability

354

implying that traffic emission profiles change during the year. PMF allocates 39%±13% of

355

Motor Vehicle Exhaust and 38%±17% to Solvent Use to toluene concentrations while our

356

study allocates 79%±19% to traffic emissions on an annual basis. For C8-aromatics, PMF

357

allocates 39%±16% to Solvent Use and 36%±13% to MVE against 72%±24% to traffic

358

emissions in our study. While traffic emissions include both exhaust and gasoline evaporative

359

losses in our study, the approach developed here shows a higher contribution of traffic source

360

to aromatic emissions for all ≥C7 aromatics than the one suggested by the PMF. Given the

361

limitations of such a comparison (see above), it is expected that the factor-of-two difference of

362

traffic is the upper limit and should be lower and within the uncertainties of both approaches.

363

Except the study by Baudic (2016), there is almost no recent study reporting on VOC source

364

apportionment in European urban areas for a direct comparison. As VOC emission regulation 16

ACS Paragon Plus Environment

Page 17 of 28

Environmental Science & Technology

365

is decided at European level, we expect our results to be extrapolated to other tertiary

366

metropolitan areas in Europe which have the same anthropogenic activity pattern. However, at

367

lower latitude cities like the ones on the Mediterranean border, higher ambient temperature

368

should favor the contribution of evaporative losses.

369

Acknowledgements

370

Nadine Locoge, Stéphane Sauvage and Antoine Waked from IMT Lille Douai (France) were

371

very helpful for consolidating VOC observations from AASQA and EMEP stations.

372

Stéphane Bonneau and Alexandre Probellet at AIRPARIF and ASPA, respectively, were

373

efficient in providing the NMHC and BTEX original data from the AASQA. The authors

374

would like to thank Roy Harrison, Dick Derwent and Sean Beevers (Kings College) for the

375

information they provided on the status of VOC emission inventories for the UK and London.

376

Thérèse Salameh benefitted from an Ile de France – R2DS postdoctoral grant (2015-2016)

377

and Anne Boynard benefitted from a financial support from CNES. The authors would like to

378

thank Cathy Clerbaux for allowing the preparation of the present work.

379

Supplement Information Available. Time series of TEX concentrations in the three cities.

380

Source contributions to TEX emissions from emission inventories. Description of the sites

381

and analytical devices for VOC measurements. Determination of the toluene-to-benzene ER

382

representative of traffic, wood combustion and gasoline evaporation emissions. Time series of

383

hydrocarbon tracer concentrations at urban and rural sites for background determination.

384

BTEX average concentrations at traffic and urban background sites and ERtraffic used in source

385

apportionment calculation. Correlation plots for the determination of ERtraffic. Source

386

contributions to tracer emissions.

387

17

ACS Paragon Plus Environment

Environmental Science & Technology

Page 18 of 28

388

References

389 390 391 392

(1) Gros, V., Gaimoz, C., Herrmann, F., Custer, T., Williams, J., Bonsang, B., Sauvage, S., Locoge, N., D'Argouges, O., Sarda-Estève, R., Sciare, J., Volatile organic compounds sources in Paris in spring 2007. Part I: qualitative analysis, Environ. Chem., 2011, Vol. 8, N°1, pp 7490

393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432

(2) Boynard, A., Borbon, A., Leonardis, T., Barletta, B., Meinardi, S., Blake, D.R., Locoge, N., Spatial and seasonal variability of measured anthropogenic non-methane hydrocarbons in urban atmospheres : Implication on emission ratios. Atmos. Environ., 82, 2014 pp.258-267. , (3) Salameh, T., Sauvage, S., Afif, C., Borbon A., Leonardis, T., Brioude, J., Waked, A., Locoge, N., Exploring the seasonal NMHC distribution in an urban area of the Middle East during ECOCEM campaigns: very high loadings dominated by local emissions and dynamics, Environ. Chem., 2015, 12(3), 316-328, http://dx.doi.org/10.1071/EN14154. (4) Dominutti P.A., Nogueira, T., Borbon, A., de Fatima Andrade, M., Fornaro, A., One-year of NMHCs hourly observations in São Paulo megacity: meteorological and traffic emissions effects in a large ethanol burning context, Atmos. Environ., 2016, 142, 371-382, http://dx.doi.org/10.1016/j.atmosenv.2016.08.008. (5) Wang, G., Cheng, S., Wei, W., Zhou, Y., Yao, S., Zhang, H., Characteristics and source apportionment of VOCs in the suburban area of Beijing, China, Air. Poll. Res. 2016, 7(4), 711-724, http://dx.doi.org/10.1016/j.apr.2016.03.006. (6) Waked, A., Sauvage, S., Borbon, A., Gauduin, J., Pallares, C., Vagnot, MP., Léonardis, T., Locoge, N., Multi-year levels and trends of non-methane hydrocarbon concentrations observed in ambient air in France, Atmos. Environ., 2016, 263-275, 141, https://doi.org/10.1016/j.atmosenv.2016.06.059. (7) Cocheo, V., Sacco, P., Boaretto, C., De Saeger, E., Perez Ballesta, P., Skov, H., Goelen, E., Gonzalez, N., Baeza Caracena, A., Urban benzene and population exposure, Nature Brief Communication, 2000, vol. 404. (8) AIRPARIF, Air quality annual report Surveillance et information sur la qualité de l’air en Ile de France en 2015, April 2016, 93 pages, http://www.airparif.asso.fr/_pdf/publications/bilan-2015.pdf (9) Carter W.P.L., Development of ozone reactivity scales for volatile organic compounds. Journal of Air and Waste Management Association, 1994, 44, pp. 881-899. (10) Carter W.P.L., Updated Maximum Incremental Reactivity scale for regulatory applications, Air Pollution Ressource Center and College of Engineering, Center for Environmental Research and Technology, University of California, Preliminary Report, 1998, Contract n° 95-308, Riverside, USA, 73 p.

18

ACS Paragon Plus Environment

Page 19 of 28

Environmental Science & Technology

433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448

(11) Deguillaume, L., Beekmann, M., Derognat, C., Uncertainty evaluation of ozone production and its sensitivity to emission changes over the Ile-de-France region during summer periods, J. Geophys. Res. Atmos., 2008, 113, 1–13, doi: 10.1029/2007JD009081.

449 450 451 452 453 454 455 456 457 458 459 460 461 462 463

(15) Ait-Helal, W., Borbon, A., Sauvage, S., de Gouw, J. A., Colomb, A., Gros, V., Freutel, F., Crippa, M., Afif, C., Baltensperger, U., Beekmann, M., Doussin, J.-F., Durand-Jolibois, R., I. Fronval, , N. Grand, N, Leonardis, T., Lopez, M., Michoud, V., Miet, K., Perrier, S., Prévôt, A. S. H., Schneider, J., Siour, G., Zapf, P., and Locoge, N., Volatile and intermediate volatility organic compounds in suburban Paris: variability, origin and importance for SOA formation, Atmos. Chem. Phys., 2014, 14, 10439–10464, doi:10.5194/acp-14-10439-2014

464 465 466 467 468 469 470 471 472 473 474 475 476

(18) von Schneidemesser, E., Monks, P. S. and Plass-Duelmer, C., Global comparison of VOC and CO observations in urban areas, Atmos. Environ., 2010, 44(39), 5053–5064, doi:10.1016/j.atmosenv.2010.09.010.

(12) Odum, J. R., Jungkamp, T. P. W., Griffin, R. J., Flagan, R. C. and Seinfeld, J. H.: The Atmospheric Aerosol-Forming Potential of Whole Gasoline Vapor, Science, 1997, 276(5309), 96–99, doi:10.1126/science.276.5309.96. (13) Ng, N. L., Kroll, J. H., Chan, A. W. H., Chabra, P. S., Flagan, R. C. and Seinfeld, J. H.: Secondary organic aerosol formation from m-xylene, toluene, and benzene, Atmos. Chem. Phys., 2007, 7(14), 3909–3922, doi:10.5194/acp-7-3909-2007. (14) Lee, A., Goldstein, A. H., Kroll, J. H., Ng, N. L., Varutbangkul, V., Flagan, R. C. and Seinfeld, J. H.: Gas-phase products and secondary aerosol yields from the photooxidation of 16 different terpenes, J. Geophys. Res., 2006, 111(D17), D17305, doi:10.1029/2006JD007050.

(16) Zhao, Y., Hennigan, C. J., May, A. A., Tkacik, D. S, de Gouw, J. A., Gilman, J. B., ,Kuster, W., Borbon, A., and Robinson, A.L., Intermediate-volatility organic compounds: a large source of secondary organic aerosol, Env. Sc. Tech., 2014, 48(23) Pages: 13743-13750, DOI: 10.1021/es5035188. (17) Molteni, U., Bianchi, F., Klein, F., El Haddad, I., Frege, C., Rossi, MJ., Dommen, J., Baltensperger, U., Formation of highly oxygenated organic molecules from aromatic compounds. Atmos. Chem. Phys. Discuss., 2016, doi: 10.5194/acp-2016-1126, under review.

(19) Warneke, C., de Gouw, J.A., Holloway, J.S., Peischl, J., Ryerson, T.B., Atlas, E., Blake, D., Trainer, M., Parrish, D.D., Multiyear trends in volatile organic compounds in Los Angeles, California: Five decades of decreasing emissions, J. Geophys. Res. Atmospheres, 2012, 117, D00V17, doi:10.1029/2012JD017899 (20) Sauvage, S., Plaisance, H., Locoge, N., Wroblewski, A., Coddeville, P., Galloo, J.C., Long term measurement and source apportionment of non-methane hydrocarbons in three French rural areas, Atmos. Environ., 2009, 43, 2430–2441.

19

ACS Paragon Plus Environment

Environmental Science & Technology

477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494

Page 20 of 28

(21) Borbon A., H. Fontaine, N. Locoge, M. Veillerot, J.C. Galloo, Developing receptororiented methods for non-methane hydrocarbon characterisation in urban air. Part II: source apportionment, Atmos. Environ., 2003, 37 4065–4076. (22) Warneke, C S., McKeen, A., de Gouw, J.A., Goldan, P.D., Kuster, W. C., Holloway, J.S., Williams, E.J., Lerner, B. M, Parrish, D. D., Trainer, M., Fehsenfeld, F.C., Kato, S., Atlas, E. L., Baker, A., and Blake, D. R., Determination of urban volatile organic compound emission ratios and comparison with an emissions database. J. Geophys. Res., 2007, 112, D10S47. http://dx.doi.org/10.1029/2006JD007930. (23) Coll, I., Rousseau, C., Barletta, B., Meinardi, S., and Blake, D.R., Evaluation of an urban NMHC emission inventory by measurements and impact on CTM results, Atmos. Env., 2010, 44(31), 3843-3855. (24) Gaimoz, C., Sauvage, S., Gros, V., Herrmann, F., Williams, J., Locoge, N., Perrussel, O., Bonsang, B., d'Argouges, O., Sarda-Estève, R., Sciare, J., 2011. Volatile Organic Compounds Sources in Paris in Spring 2007. Part II: Source Apportionment Using Positive Matrix Factorisation. Environ. Chem., 2011, 8, no. 1, 91–103.

495 496 497 498 499 500 501 502 503 504 505 506 507

(25) Salameh, T., Borbon, A., Afif, C., Sauvage, S., Leonardis, T., Gaimoz, C., and Locoge, N.: Composition of gaseous organic carbon during ECOCEM in Beirut, Lebanon: new observational constraints for VOC anthropogenic emission evaluation in the Middle East, Atmos. Chem. Phys., 2017, 17, 193-209, doi:10.5194/acp-17-193-2017, 2017.

508 509 510 511 512 513 514 515 516

(28) Baudic, A., Gros, V., Sauvage, S., Locoge, N., Sanchez, O., Sarda-Estève, R., Kalogridis, C., Petit, J.-E., Bonnaire, N., Baisnée, D., Favez, O., Albinet, A., Sciare, J., and Bonsang, B.: Seasonal variability and source apportionment of volatile organic compounds (VOCs) in the Paris megacity (France), Atmos. Chem. Phys., 2016, 16, 11961-11989, doi:10.5194/acp-1611961-2016

517 518 519 520

(30) Borbon, A., Gilman, J.B., Kuster, W. C., Grand, N., Chevaillier, S., Colomb, A., Dolgorouky,C., Gros, V., Lopez, M., Sarda-Esteve, R., Holloway, J., Stutz, J., Perrussel, O., Petetin, H., McKeen, S., Beekmann, M., Warneke, C., Parrish, D.D., and de Gouw, J., Emission ratios of anthropogenic VOC in northern mid-latitude megacities: observations vs.

(26) Niedojadlo, A., Becker, K. H., Kurtenbach, R., Wiesen, P., The contribution of traffic and solvent use to the total NMVOC emission in a German city derived from measurements and CMB modelling, Atmos. Environ., 2007, 41, 7108. doi:10.1016/J.ATMOSENV. (27) Lanz, V. A., Hueglin, C., Buchmann, B., Hill, M., Locher, R., Staehelin, J. and Reimann, S.: Receptor modeling of C2–C7 hydrocarbon sources at an urban background site in Zurich, Switzerland: changes between 1993–1994 and 2005–2006, Atmos. Chem. Phys., 2008, 8(9), 2313–2332, doi:10.5194/acp-8-2313-2008.

(29) Dollard, G. J., Dumitrean, P., Telling, S., Dixon, J. and Derwent, R. G., Observed trends in ambient concentrations of C2–C8 hydrocarbons in the United Kingdom over the period from 1993 to 2004, Atmos. Environ., 2007, 41, 2559–2569.

20

ACS Paragon Plus Environment

Page 21 of 28

521 522 523 524 525 526 527 528 529 530 531

Environmental Science & Technology

emission inventories in Los Angeles and Paris. J. Geophys. Res., 2013, vol. 118, 2041–2057, doi:10.1002/jgrd.50059 (31) Yuan, B., Shao, M., Lu, S., Wang, B., Source profiles of volatile organic compounds associated with solvent use in Beijing, China, Atmos. Environ., 2010, 44, 1919-1926. (32) Salameh, T., Afif, C., Sauvage, S., Borbon, A., and Locoge, N.: Speciation of NonMethane Hydrocarbons (NMHC) from anthropogenic sources in Beirut, Lebanon, Environ. Sci. Pollut. Res., 2014, 21, 10867–10877, doi:10.1007/s11356-014-2978-5.

21

ACS Paragon Plus Environment

Environmental Science & Technology

532

Page 22 of 28

Figures and Tables

533 534

535 536 537 538 539 540 541 542 543 544 545

Figure 1 : Time series of toluene monthly mean concentrations in Paris, Strasbourg and London at traffic and urban background sites since 2000 (left panel). The grey shaded area corresponds to the standard deviation of the monthly mean concentrations at the background sites. Associated apportioned emissions of toluene extracted from local emission inventories are reported on right panel (grey: road transport ; red: solvent use or residential for Strasbourg; dark blue: extraction and distribution of fossil fuel; yellow: non-industrial combustion and industrial activities for Strasbourg; light blue: other activities.

22

ACS Paragon Plus Environment

Page 23 of 28

Environmental Science & Technology

546

a)

547

548 549 550 551 552 553 554 555

b)

Figure 2: (a) Scatterplots of toluene versus benzene concentrations in January (blue) and July (red) months at London, Paris and Strasbourg urban background sites (2008-2013). Dotted lines are average toluene-to-benzene ERs representative of traffic (Aït-Helal et al., 2015), domestic wood combustion (Gaeggeller et al. 2008; Evtyugina et al., 2014; Baudic et al., 2016) and gasoline evaporation emissions (Salameh et al., 2017). (b) Time series of monthly toluene-to-benzene urban enchancement ratios (ER) derived from the slopes of correlation plots for Paris and Strasbourg at traffic and urban background sites

556

23

ACS Paragon Plus Environment

Environmental Science & Technology

Page 24 of 28

557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604

Figure 3 : Time series of toluene concentration terms in Paris, Strasbourg and London between 2008 and 2013. Data are smoothed a 30-day basis. Below time series, pie charts represent the average summertime (JJA) percent contribution of the traffic term (in grey), 24

ACS Paragon Plus Environment

Page 25 of 28

605 606 607

Environmental Science & Technology

background term (in blue) and non-traffic term (in red) to toluene concentrations. Framed numbers are the annual average contribution of traffic and its standard deviation.

25

ACS Paragon Plus Environment

Environmental Science & Technology

608 609

Page 26 of 28

Table 1: VOC data used for this study Location

Site

Type

Coord.

Network

Period

% missing data

Reference

London

Marylebone Road Eltham

Traffic Urban

51.52 N, -0.1546 W 51.45 N, 0.071 E

Hydrocarbon Automatic Network Hydrocarbon Automatic Network

2008 - 2013 2008 - 2013

23% 19%

Dollard et al., 2007 Dollard et al., 2007

Paris

Auteuil Les Halles - Crillon Guy Moquet tunnel

Traffic Urban Tunnel

48.85N 2.25E 48.86N 2.34 E 48.01N 2.65E

AIRPARIF AIRPARIF PREQUALIF experiment

2008 - 2011 24% 2008 - 2013 6% Sept-Oct. 2012 2%

Waked et al., 2016 Baudic et al., 2016

Clémenceau Strasbourg Ouest

Traffic Urban

48.35N 7.44E 48.36N 7.42E

ASPA ASPA

2008 - 2013 2008 - 2013

Boynard et al., 2014

Strasbourg

3% 16%

610 611 612 613 614

26

ACS Paragon Plus Environment

Page 27 of 28

Environmental Science & Technology

615

Aromatics

Cities

2008

2009

2010

2011

2012

2013

Paris

(m+p)-xylenes

London

Strasbourg

Paris

o-xylene

London

Strasbourg

Paris

ethylbenzene

London

Strasbourg

27

ACS Paragon Plus Environment

Environmental Science & Technology

Page 28 of 28

616

Table 2:

617 618 619 620 621 622

Table 2: Mean contribution in % of traffic emissions (in grey) to (m+p)-xylenes, o-xylene and ethylbenzene concentrations in summer (JJA) in Paris, London and Strasbourg between 2008 and 2013. Non-traffic contribution (in red) includes the background. Framed numbers represent the annual mean contribution of traffic emissions

28

ACS Paragon Plus Environment